UC Irvine UC Irvine Previously Published Works

Title Histocompatibility bioassays of population structure in marine : Clonal structure in verongia longissima and lotrochota birotulata

Permalink https://escholarship.org/uc/item/0g93w1qg

Journal Journal of Heredity, 74(3)

ISSN 0022-1503

Authors Neigel, JE Avise, JC

Publication Date 1983

DOI 10.1093/oxfordjournals.jhered.a109750

License https://creativecommons.org/licenses/by/4.0/ 4.0

Peer reviewed

eScholarship.org Powered by the California Digital Library University of California The Journal of Heredity 74: 134-140. 1983.

Histocompatibility bioassays of population structure in marine sponges

Clonal structure in Verongia longissima and lotrochota birotulata

ABSTRACT: Clonal population structure In two marine sponges, Verongia longissima and lotrochota birotulata, was examined with a self-recognition bioassay. The bloassay con- sists of grafts of branch segments between conspecific individuals. Results were consis- tent with the operational properties expected of a precise histocompatibility system. Auto- grafts exhibited acceptance responses; grafts between individuals separated by large dis- tances exhibited rejection responses; individuals were not limited to a single mode of re- sponse at one time; and all identity relationships were transitive. Clonal population struc- ture was assessed by examining the relationship between graft response and donor-to- reclplent distance, and by actually mapping the distributions of particular clones. Clones Joseph E. Neigel of lotrochota birotulata were usually restricted to single coral heads or small patch reefs (1-3 m diameter). For Verongia longissima, which can grow directly upon the coral rubble John C. Avise surrounding coral heads and patch reefs, individual clones often occupied larger areas (up to 10 m diameter). The spatial patterns of clonal distributions are readily interpreted as consequences of the particular demographies and habitat specificities of these two species.

THE SPATIAL distribution of asexual linea- ena as bioassays of clonal population structure ges or clones is an important component of has been demonstrated with self-incompati- genetic structure in the populations of many bility in clover and grasses12"14, histocompa- plants and sedentary invertebrates. The tibility in parthenogenetic vertebratesl-6-28, preservation of characteristic genotypes by and interclonal aggression in sea anemones35. clonal propagation implies that clonal struc- Of potentially much wider application, how- ture can be resolved by any method that assays ever, are invertebrate "histocompatibility" a sufficiently polymorphic subset of the ge- systems, which recently have been demon- nome. Polymorphic allozyme and morpho- strated in several major phyla5'19-20^2'24'25. If logical markers have been thus employed to invertebrate histocompatibility systems prove identify clones within natural popula- to be generally capable of resolving genetic tions 17,21,31.32,36,37 However, another class of variation in natural populations, they will then genetic markers is perhaps ideally suited for constitute a powerful tool for analysis of pop- the analysis of clonal population structure: ulation structure in a large and diverse group those manifested in various self-recognition of organisms. The authors arc, respectively, graduate student and phenomena. associate professor in the Department of Molecular and From what is already known of invertebrate Population Genetics, the University of Georgia, Athens, Self-recognition phenomena, such as self- histocompatibility systems and histocompa- GA 30602. They wish to thank the University of the incompatibility in flowering plants, and his- tibility systems in general, it is apparent that West Indies' Discovery Bay Marine Laboratory in Ja- not all responses in contacts between conspe- maica for their cooperation in the field. This work was tocompatibility in , are often under the supported by NSF Doctoral Dissertation Improvement control of highly polymorphic genetic systems, cific individuals are mediated by precise self- Grant DEB-8I-136I7 to J. Neigel and J. Avise, and which effectively provide each sexually derived recognition systems4-7'29. Furthermore, the NSF Grant DEB-8O-22I35 to J. Avise. Contribution individual and its clonal descendants in the resolving power of a self-recognition system, No. 251 from the Discovery Bay Marine Laboratory, 2 18 23 26 University of the West Indies. population with a unique label ' - ' . The which is presumably a function of inter-clonal © 1983, American Genetic Association. possibility of using self-recognition phenom- genetic differences, may vary among popula-

134 tions. It is therefore necessary to test a self- recognition assay in the population in which it is being applied before making a definitive evaluation of its results. Recently we suggested a protocol for testing any self-recognition bioassay of clonal identity30. For a true self- recognition phenomenon that distinguishes among clonal genotypes, it is predicted that specific operational properties must charac- terize the behavior of the phenomenon with ' respect to genetic relationships: individuals should be capable of responding selectively to specific contacts; interactions between parts of the same individual or between members of the same clone should always elicit self re- FIGURE 3 Iotrochota birotulata at the EBR site. This individual is attached to a non-living coral head, sponses; interactions with different clones FIGURE 1 Verongia longissima at the WFR site of Discovery Bay. The runners of this individual are and is isolated from other areas of suitable substrate should always be responded to as non-self; extended over a coral head. by sand. responses should be reproducible; and rela- tionships defined by self responses should be transitive. These operational properties were demonstrated for naturally occurring and bioassay approach to population structure rubble. Some individuals appeared to be ac- experimental histocompatibility contact re- analysis. tively spreading over the substratum, with sponses in populations of the reef-building their encrusting bases and branches inter- t coral Acropora cervicornis. The histocompa- connected by runners (Figure 1). Smaller Materials and Methods tibility bioassay was then used to describe specimens possessed few or no branches. Our clonal structure in Acropora*0. The field work was conducted at the Dis- study site for this species was on the West Fore covery Bay Marine Laboratory of the Uni- Reef (WFR) of Discovery Bay, at a depth of In the present study we have extended the versity of the West Indies, Jamaica in March about 15 m (Figure 2). Kaye and Ortiz25 have use of histocompatibility analysis of clonal and April 1982. The two species of previously shown, for a Barbados population population structure to the most primitive studied are both common on Carribean coral of this species, that both "acceptance" and invertebrate phylum, the sponges (Porifera). reefs at depths less than 20 meters. Verongia "rejection" responses could be obtained in Certain biological characteristics of sponges, longissima (Carter) was found in Jamaica on grafts between individuals. which they share with many other sessile hard substrata in non-cryptic reef locations. Iotrochota birotulata (Higgin), like V. marine invertebrates, make them especially The typical growth form was a set of several longissima, also is found on hard substratum well suited for histocompatibility analysis. branches about lem in diameter, repent or in open areas, except that it is generally con- Vegetative propagation is widespread in erect, and originating from an encrusting base fined to more solid coral heads or patch reefs; sponges, presumably because of their primitive attached to either a coral head or loose coral it is not found on loose rubble or sand. The organization and regenerative capabilities. typical growth form was a bushy clump of > Sexual reproduction in most sponges involves erect or sprawling branches of lem to 10 cm the production of motile larvae, which settle thickness (Figure 3). This species was studied on suitable substrata, metamorphose into ju- at two sites in Discovery Bay: on the West veniles and become morphologically indis- Fore Reef, in the same locale where V. long- tinguishable from vegetatively derived indi- issima was studied, and at an East Back Reef viduals3. Clones in sponge populations there- (EBR) site where a dense population was fore are initiated by larval recruitment and found in water 2-10 m deep (Figure 2). spread within the habitat by vegetative prop- The question of what the term "individual" agation. Finally, although sponges are char- should refer to when considering colonial an- acterized by a primitive cellular level of or- imals has long been deliberated27, and is ganization, some have been reported to possess especially acute in the case of sponges, where a capacity for precise self vs. non-self dis- the usual structural criteria for the individual crimination that outwardly resembles verte- units of colonial organization are inappro- > brate histocompatibility19'25. priate10. Here we will follow the usage of In this paper we will 1) consider the opera- Hartman and Reiswig16, which avoids much tional properties exhibited by histocompati- of this ambiguity, and consider "entire con- bility responses in populations of two marine fluent specimens" as individuals. demospongia, Verongia longissima and Io- Grafting experiments were performed in trochota birotulata; 2) examine the spatial situ with the use of SCUBA. Branch segments distributions of clones defined by histocom- 3-10 cm long for V. longissima or 10-25 cm patibility assays; 3) discuss clonal structure in long for /. birotulata were tied with nylon FIGURE 2 Map of Discovery Bay, Jamaica, in- populations of these sponges in relation to their dicating sites where sponge grafting experiments were monofilament to a branch on the same or dif- utilization of the reef habitat; and 4) evaluate conducted. WFR, west fore reef site; EBR, east back ferent individual. Each graft was identified by the general effectiveness of the self-recognition reef site. a plastic tag labeled with a set of punched

Neigel and Avise: Clonal structure in sponge populations May/June 1983 135 B

FIGURE 4 Graft responses of Verongia longissima. A—acceptance response. B—rejection response.

holes. Scoring of graft responses was essen- results were entirely consistent with the initial individual responded differentially to tially "blind" with respect to expected out- scorings. Acceptance and rejection responses, grafts received from different donors. This comes, because each graft was scored prior to and the histology of graft interfaces have been demonstrates the necessary capacity for examination of its identification tag. Each described by Kaye and Ortiz25. specific responses to individual donors, graft was scored independently by both JN For /. birotulata, a total of 285 grafts was rather than a genera! response to all grafts and JA. performed (26 at the WFR site, and 259 at the by an individual sponge. After some preliminary experiments, it was EBR site). Among the total, 270 (95 percent) 2) Self-grafts exhibit acceptance re- found that grafts of V. longissima could be were subsequently located and scored after sponses. A total of 13 autografts for V. scored within one week, and grafts of /. bir- 5-8 days. Acceptance responses exhibited longissima and 37 for /. birotulata were otulata within 5 days, provided a large enough rapid fusion of the graft and host branches to made; the procedure involved grafting a area of contact (>2 cm2) was present at the the extent that their contours merged and no branch segment to a new location on the graft interface. Grafts that had become loose trace of the original graft interface could be same individual. In all cases the expected were retied and later rescored. Subsets of detected. A subset of 44 accepting grafts re- acceptance response was observed. grafts for both species were rescored at later scored after 9-16 days maintained this ap- 3) Grafts between individuals separated times to confirm that initial scorings were re- pearance. Rejection responses were charac- by large distances exhibit rejection re- liable. terized only by a lack of fusion; cuticle for- sponses. For the species investigated, mation, tissue necrosis, or other active mani- which accomplish vegetative dispersal by festations of a rejection were not observed growth over the substratum, it seemed Results (Figure 5). After 9-16 days, a subset of 78 of reasonable to assume that the spatial ex- Among a total of 181 V. longissima grafts these rejecting grafts had all remained un- tent of a clone would be limited. There- fused, although in every case an adequate area fore, if histocompatibility responses were performed, 177 (98 percent) were subse- 2 quently located and scored after 5-20 days. of tissue contact (>2 cm ) was present at the precise, the frequency of graft acceptance All grafts were viable, and exhibited either of graft interface. between individuals should decrease with two characteristic responses—acceptance or the distance between them, and should rejection. Acceptance responses involved the eventually approach zero. This expected fusion of the grafted branch segment to the Opera tioaal properties trend was evident for both species (Figure recipient branch, so that an unbroken surface Grafting experiments with Verongia long- 6). For V. longissima, all 68 grafts be- was formed at the graft interface. Rejection issima and Iotrochota birotulata demon- tween individuals greater than 10 m apart responses were initially manifested as a lack strated the following operational properties: exhibited rejection responses. All 105 of fusion at the graft interface; however, after 1) An individual is not limited to a single grafts between /. birotulata individuals several weeks a cuticle developed between the mode of response at one time. Thirteen greater than 2.7 m apart resulted in rejec- donor and recipient branches (Figure 4). Ten cases for V. longissima and 24 cases for /. tions. These results indicate that interclo- grafts were rescored after 7-29 days, and the birotulata were observed where a single nal grafts within a population can be accu-

136 The Journal of Heredity Neigel and Avise: Clonal structure in sponge populations FIGURE 5 Graft responses of Iotrochota birotulala. A—acceptance response. B—rejection response.

rately recognized as non-self, and re- pair were related by an acceptance re- the pairs with the smallest donor-to-recipient jected. sponse were "predictive," in the sense that distances. Five distinct clones of/, birotulata 4) Individuals related by graft accept- one relationship could be predicted from were identified in a set of six individuals within ances respond alike to grafts from other the other two by hypothesizing transitiv- a 75 m2 area on the WFR. In this set, the two individuals. If graft acceptance implies ity. No intransitive relationships were ob- individuals identified as clonemates were at- genetic identity, putative clonemates served. tached to the same coral head. Each member should exhibit parallel responses in grafts of a set of seven /. birotulata individuals Clonal population structure with other individuals; true identity rela- within a 300 m2 area at the EBR site displayed tionships are necessarily transitive. If For vegetatively propagating organisms, a unique clonal identity. graft responses were either influenced by like sponges, the spatial extent of clones ulti- Histocompatibility assays also were used to non-genetic factors, or were under the mately should be limited by their ability to provide direct representations of clonal pop- control of an imprecise genetic system, in- disperse. As described in the previous section, ulation structure by actually mapping the transitivities could occur. In Table I are to provide a test of our bioassay's accuracy we spatial distributions of individual clones. presented, for both sponge species, the re- assumed the validity of this proposition, and sults of all sets of grafts that defined rela- required that grafts between individuals sep- tionships among three or more individu- arated by large distances exhibit rejection als. Relationships in which at least one responses. More explicit assessments of the Table I. Grafting assay defined relationships among sets of three or more individual sponges; relationship between donor-to-recipient dis- values represent number of observations tance and frequency of graft rejections can be interpreted as descriptions of clonal population Relationship V. longissima /. birotulata structure. A comparison of the two species in Transitive our study on this basis (Figure 6), indicates A -B that clones of V. longissima tend to be of A = C 10 5 greater spatial extent than clones of /. bir- B = c otulata. A - B The number of clones represented in a set A 5* C 24 12 of individuals can be determined exactly if all B *• C pairwise relationships among those individuals Intransitive are known. All possible pairwise combinations A = B O.OI-0J OJ-I I—10 10-100 of grafts between individuals (not including A = C 0 0 DONOR TO RECIHHT DISTANCE (UCTUS) reciprocal grafts) were tested for three sets of B *C sponges. Eight distinct clones of V. longissima Nonpredictive FIGURE 6 Frequency of graft rejection in Ver- were identified in a set of 10 individuals within 7— MJ ongia longissima and Iotrochola birotulata for five A a 1800 m2 area on the WFR. The two pairs of Hi 51 donor-to-recipient distance classes, with 95 percent B confidence intervals. individuals identified as clonemates also were

Neigel and Avise: Clonal structure in sponge populations May / June 1983 137 FIGURE 8 Map of Verongia longissima at the WFR site (LTS Reef). Solid drdes, triangles, and square indicate the specific donal identities determined for individuals at those positions; open drdes indicate individuals located in the mapped area, but not used in graft comparisons.

FIGURE 7 Map of a single clone of Verongia longissima at the WFR site (Danring Lady Reef). Solid rircles indicate positions of donemates; open drcles, non-clonemates. mapped V. longissima clones is the homoge- tensive genetic polymorphisms that can serve neity of their distributions; boundaries could as clone-specific markers. The possibility of be defined that included all the members of a such an approach had not been considered specific clone but did not include any members until recently. It generally had been believed Particular clones of V. longissima and /. bir- of other clones. that a system for precise self vs. non-self dis- otulata were mapped by selecting individuals crimination was a special feature of vertebrate Habitat correlation in an area, recording their positions, and evolution. The complexity of the vertebrate grafting them with a standard individual. The There were no obvious features of the hab- immune system, with its architecture of spe- V. longissima clone appeared as a cluster of itat that corresponded with the distributions cialized cells, tissues and organs, seemed to individuals surrounded by non-clonemates of V. longissima clones. However, an inter- preclude a pre-vertebrate origin8'9'". Results (Figure 7). The map of the /. birotulata clone esting relationship between clonal distribution of early transplantation studies generally up- at the EBR site reflected the general pattern and habitat structure is suggested by the /. held this view; lower invertebrate hosts ap- of high clonal diversity exhibited by this birotulata grafting data. Clones appeared to peared to accept any physiologically compat- species; no acceptance responses were found be restricted to single coral heads or solid ible tissue, even across species boundaries4'29. •> in grafts with the tested individuals. For V. patch reefs, the discrete "patches" of the However, a search for the pre-immunocom- longissima, an additional mapping approach habitat utilized by this species. Grafts were petent stage of metazoan evolution among was used, which made use of the rapidity of the made between individuals on separate habitat living representatives of the invertebrates re- graft responses and the operational property patches, and between individuals occurring cently led Hildemann and co-workers19-20 to of transitivity. An area containing a dense within the same patch. The graft responses conclude that histocompatibility systems, concentration of V. longissima was selected and donor-to-recipient distances for these characterized by precision and immunological and grafts among the sponges were made, experiments are shown graphically in Figure memory, could be found in even the lowest initially at random. Scoring the initial grafts 9. There were only three examples, out of 127 multicellular animals. revealed groups of compatible individuals, graft comparisons between individuals on To reconcile the apparent contradiction which were then extended to include addi- separate habitat patches, of a clone being between the early findings of indiscriminant tional individuals and merged into larger represented in two rather than one habitat compatibility, and some more recent findings groups by the application of further grafting patch. However, /. birotulata graft responses of precise allogcncic incompatibility, it must comparisons. Many relationships were inde- also showed a parallel dependence on distance be concluded that either the choice of organ- 4 pendently verified by multiple graft compar- alone, both for within-patch and between- isms has been critical in these studies, or al- isons. A total of 44 grafts was made within this patch comparisons. There were no apparent ternatively, differences in technique and set, which unambiguously assigned each of the differences in /. birotulata clonal population methodology have played a larger role then is 35 grafted individuals to one of three clones. structure between the WFR and EBR sites. generally suspected. Of the latter, it may be An additional 53 individuals were located in significant that precise allogeneic graft dis- -* this area; these were tentatively assigned to crimination has been consistently found when one of the three clones on the basis of their Discussion whole sponges, or segments with their outer * proximity to individuals of known clonal pinacodermal covering still intact, are placed identity. The resulting multi-clone map is Self-recognition phenomena in lower ani- in apposition19-25 (and this study); and where shown in Figure 8. Characteristic of all four mals, and histocompatibility in particular, may provide a general means of detecting ex- lack of discrimination, or ambiguous results

138 The Journal of Heredity Neigel and Avise: Clonal structure in sponge populations 25 -r BETVEEN PATCHES requires not only the existence of the biological self-recognition system, but also a method of ACCEPTANCES extracting information from it accurately. The approach is limited only by the availability of a set of known genetic relationships to serve as controls, and the assumption that they are representative of the unknown relationships 0 J- being determined. We have introduced one method of testing the assumption that histo- 2S -r 0.01 • . e.i I m. 10 100 compatibility responses in the populations we have studied accurately reflect clonal rela- REJECTIONS tionships. By examining assay-defined rela- tionships among sets of three or more indi- viduals, identity relationships can be charac- terized as either transitive or intransitive. 0 J- , II III I Because in our study identity relationships were always transitive, we conclude that: 1) grafting responses were reproducible (alter- native outcomes would have produced in- transitivities), and 2) relationships defined by graft compatibility fit a model in which com- plete matching of histocompatibility genotypes is required for a graft acceptance. Thus, there was no evidence for a threshold model in which graft acceptances could occur when histo- compatibility genotypes were similar, but not 25 T WITHIN PATCHES identical. This is an important result, because ACCEPTANCES a threshold system of histocompatiblity has been demonstrated for at least one inverte- brate, the colonial ascidian, Botryllus schlosseri34. In the Botryllus system, one locus with many alleles determines compati- 0 -L I ,1 I .III ll ,, , bility. If two individuals share either one or both alleles their tissues are compatible (a one-allele threshold). Thus, all parent-off- 25 -r a.81 ». 0.1 10 100 spring and many sibling relationships are compatible. Intransitive compatibilities would REJECTIONS be observed among Botryllus colonies with the allelic compositions AB, BC, and CD. The population genetic structures evident from spatial distributions of histocompatible e J- I I I i I I individuals of V. longissima and /. birotulata can be readily explained in terms of clonal FIGURE 9 Graft responses and donor-to-recipient distances (logarithmic scale) for grafts between Iotrochota propagation. This interpretation is consistent birotulaia from different patches of suitable substrate, and between individuals from the same patches. with the results of our experimental grafting studies, and with our observations of individ- uals in the process of extending branches or runners across the substrate. However, sexual reproduction potentially could have a similar have been obtained, the technique has involved pinacoderm were inserted subdermally into effect on genetic population structure, if the subdermal insertion of a "core" excised hosts of the same species, we could not observe gamete and larval dispersal were so limited as ^ from the donor, through an incision made in any distinct responses that would distinguish to almost completely restrict gene flow within 33 38 the host7'29. The potential importance of allografts from autografts (unpublished ob- a population - . Inbred patches of closely technique became apparent in our study with servations). For /. birotulata, graft rejections related individuals, monomorphic for the ge- Iotrochota birotulala. The branch segments could only be detected as a passive lack of fu- netic determinants of histocompatibility, we ultimately used as grafts were as large as sion between apposed pinacoderms. would appear as single clones. This caveat is many of the individuals in the population, and By defining the behavior of a self-recogni- shared by all methods, including morpholog- so the distinction between "donor" and "re- tion phenomenon in terms of operational ical and electrophoretic methods, that examine r cipient" has simply been used as a convenience properties, experimental methods are included a subset of the genome to characterize clones. to describe the final location of a pair of as part of the phenomenon. This convention We have assumed that gene flow within the grafted individuals. In preliminary experi- can be used to address the suitability of the populations studied was sufficient to prevent ments, where small explants scraped free of phenomenon as a workable bioassay, which extreme inbreeding effects.

Neigel and Avise: Clonal structure in sponge populations May /June 1983 139 The results of the present study suggest one 3. BERGQUIST, P. R. Sponges. U. of Calif. Press, 22. IVKER, F. B. A hierarchy of histocompatibility in way in which microgeographic habitat varia- Berkeley. 1978. Hydractinla echinata. Biol. Bull. 143:162-174. 4. CAMPBELL, R. D. and C. BIBB. Transplantation 1972. tion may influence clonal distributions in in coelenterates. Transplant. Proc. 2:202-211. 23. KALLMAN, K. D. An estimate of the number of populations of Iotrochota birotulata. The 1970. histocompatibility loci in the teleost Xiphophorus vegetative extension of /. birotulata is ulti- 5. COOPER, E. L. Transplantation immunity in hel- maculatus. Genetics 50:583-595. 1964. minths and annelids. Transplant. Proc. 2:216-221. 24. KARP, R. D. and W. H. HiLDEMANN. Specific mately constrained by the spatial limits of 1970. allograft reactivity in the sea star Dermasterias coral heads and solid patch reefs upon which 6. CUELLAR, O. Genetic homogeneity and spedatkm imbricata. Transplantation 22:434-439. 1976. it can attach. Small distances (up to approxi- in the parthenogenetic lizards Cnemidophorus 25. KAYE, H. and T. ORTIZ. Strain specificity in a mately 1 m) across stretches of unsuitable velox and C. neomexicanus: evidence from in- tropical marine sponge. Marine Biol. 63:165-173. traspecific histocompatibility. Evolution 31:24-31, 1981. substratum can be traversed by extended 1977. 26. KLEIN, J. Genetic polymorphism of the histo- branches (cover photo), while greater dis- 7. CURTIS, A. S. G., J. KERR, and N. KNOWLTON. compatibility-2 loci of the mouse. Ann. Rev. Genet. tances over sand or loose coral rubble probably Graft rejection in sponges: genetic structure of 8:63-77. 1974. accepting and rejecting populations. Transplan- 27. MACKIE, G. O. Siphonophores, bud colonies, and act as barriers to vegetative expansion. The tation33:127-133. 1982. superorganisms. In The Lower Metazoa; Com- distributions of/, birotulata clones appear to 8. CUSHING, J. E. Invertebrates, immunology and parative Biology and Physiology. E. C. Dougherty, reflect this limitation on vegetative propaga- evolution. AS.E.B. Fed. Proc. 26:1666-1670. Ed. Univ. Calif. Press, Berkeley, p. 329-337. tion. Clones were restricted to single coral 1967. 1963. 9. FAVOR, C. B. Comparative immunology and the 28. MASLIN, T. P. Skin grafting in the bisexual Teeid heads or patch reefs, except for two instances phylogeny of homotransplantation. Ann. N.Y. lizard Cnemidophorous sexlineatus and in the in which the distance between patches occu- Acad. Sci. 73:590-598. 1958. unisexual C. tesselatus. J. Exp. Zool. 166:137- pied by a clone were within the range an ex- 10. FRY, W. G. , the individual and the 150. 1967. sponge. Syst. Assoc. Special 11:49-80. 1979. 29. MOSCONA, A. A. Cell aggregation: properties of tended branch could traverse. In contrast with 11. GOOD, R. A. and B. W. PAPERMASTER. Ontog- specific cell-ligands and their role in the formation /. birotulata, V. longissima can utilize loose eny and phylogeny of adaptive immunity. Adv. of multicellular systems. Dev. Biol. 18:250-277. coral rubble as a substratum for attachment. Immunol. 4:1-115. 1964. 1968. The more extensive clonal distributions of this 12. HARBERD, D. J. Observations on natural clones 30. NEIGEL, J. E. and J. C. AVISE. Clonal structure of Trifolium repens L. New Phytol. 62:198-205. in natural populations of the reef-building coral species did not appear to have been molded by 1962. Acropora cervicornis: self-recognition analysis and any recognizable features of the habitat. In a 13. Some observations on natural clones in demographic interpretation. Evolution in press. previous study of clonal structure in popula- Festuca ocina. New Phytol. 61:85-100. 1962. 1983. 14. and M. OWEN. Some experimental obser- tions of the reef-building coral, Acropora 31. OINONEN, E. Sporal regeneration of bracken in vations on clone structure of a natural population Finland in the light of the dimensions and age of its cervicornis, simple models incorporating of Festuca rubra. New Phytol. 6853-104.1969. clones. Ada Forest. Fenn. 83:3-96. 1967. empirically estimated rates of vegetative 15. HARPER, J. L. The demography of plants with clonal growth. In Structure and Functioning of 32. PARKER, E. D., SR. Phenotypic consequences of propagation, mortality, and sexual recruit- Plant Populations. A. H. Freyson and J. W. Wol- parthenogenesis in Cnemidophorus lizards. I. ment could account for both qualitative and dendorp, Eds. North Holland, Amsterdam, p. Variability in parthenogenetic and sexual popu- 27-48. 1978. lations. Evolution 33:1150-1165. 1979. quantitative features of the observed clonal 33. ROHLF, F. J. and G. D. SCHNELL. An investiga- distributions30. These examples suggest that 16. HARTMAN, W. D. and H. N. REISWIG. The in- dividuality of sponges. In Colonies. R. S. tion of the isolation-by-distance model. Am. Nat. a detailed consideration of basic demographic Boardman, A. H. Cheetham, and W. A. Oliver, Jr., 105:295-324.1971. processes may provide immediate explanations Eds. Dowden, Hutchinson and Ross, Stroudsberg. 34. SCOFIELD, V. L, J. M. SCHLUMPBERGER, L. A. WEST, and I. L. WEISMAN. Protochordate allo- for observed patterns of clonal structure. The p. 567-584. 1973. 17. HEBERT, P. D. N. Enzyme variability in natural recognition is controlled by a MHC-like gene sys- use of histocompatibility and other self-rec- populations of Daphnla magna I. Population tem. Nature 295:499-502. 1982. ognition phenomena to determine clonal structure in East Anglia. Evolution 28:546-556. 35. §EBENS, K. P. Asexual reproduction in Antho- structure in populations of lower animals can 1975. pleura elegantissima (Anthozoa: Actiniaria): 18. HiLDEMANN, W. H. and R. HAAS. Histocom- seasonality and spatial extent of clones. Ecology provide many opportunities for analyses of patibility genetics of bullfrog populations. Evolu- 63:434-444. 1982. these systems. tion 15:267-271. 1961. 36. SHICK, J. M. and A. N. LAMB. Asexual repro- 19. , I. S. JOHNSTON, and P. L. JOKIEL. Im- duction and genetic population structure in the munocompetence in the lowest metazoan phylum: colonizing Haliplanella luciae. Biol. transplantation immunity in sponges. Science Bull. 153:604-617. 1977. References 204:420-422. 1979. 37. SILANDER, J. A. Microevolution and clone 20. , R. L. RAISON, G. CHEUNG, C. J. HULL, structure in Spartina patens. Science 203:658-660. 1. ANGUS, R. A. and R. J. SCHULTZ. Clonal diver- L. AKAKA, and J. OKAMOTO. Immunological 1979. sity in the unisexual fish Poeciliopsis monacha- specificity and memory in a scleractinian coral. 38. TURNER, M. E., J. C. STEPHENS, and W. W. ludda: a tissue graft analysis. Evolution 33:27-40. Nature 270:219-223. 1977. ANDERSON. Homozygosity and patch structure 1979. 21. HOFFMANN, R. J. Genetics and asexual repro- in plant populations as a result of nearest neighbor 2. BATEMAN, A. J. Number of S-alleles in a popu- duction of the sea anemone Metridium senile. Biol. pollination. Proc. Natl. Acad. Sci. 79:203-207. lation. Nature 160:337. 1947. Bull. 151:478-488. 1976. 1982.

140 The Journal of Heredity Neigel and Avise: Clonal structure in sponge populations